A redated post.
Keith Parsons thinks the argument is not so overrated. See these comments in a two-journal exchange with me:
It will simply not do far Hart (or Reppert) to take refuge in familiar Humean conundrums about causality. Much of the progress of science has been progress in understanding how things interact: Plate Tectonics tells us how crustal plates interact to produce earthquakes, volcanoes, mountains and other geological phenomena. Likewise, ecology helps us to understand the enormously complex interactions with their physical environment and each other. Molecular biology explains the interactions of complex molecules, e. g. enzymes and their substrates. Even at the rock-bottom level of quarks and gluons we have well-confirmed, mathematically precise theories that often make (as in Quantum Electrodynamics) astonishingly accurate predictions. These theories tell us how fundamental particles interact.
Keith M. Parsons, "Further Reflections on the Argument from Reason" Philo (Spring-Summer 2000), 93.
As Lycan notes above, Descartes took this objection seriously, and he should have. Surely dualists owe the rest of us some sort of account. After all they posit and entity that has no physical properties (and consequently is undetectable by any empirical means), but which is not an abstract entity since it somehow interacts with physical things—in a way that violates conservation laws, by the way. Souls could not have been produced by physical means, and their putative existence raises a host of unanswerable questions. (For example, at what point at the evolution of hominids did our ancestors acquire souls? Homo habilis or Homo erectus, maybe).
Keith M. Parsons, "Need Reasons be Causes? A Further Reply to Victor Reppert's Argument from Reason", Philosophia Christi (Vol. 5, No. 1, 2003) pp. 72-73.